Sei sulla pagina 1di 10

Preparative Biochemistry and Biotechnology

ISSN: 1082-6068 (Print) 1532-2297 (Online) Journal homepage: https://www.tandfonline.com/loi/lpbb20

Optimization of non-detergent treatment for


enveloped virus inactivation using the Taguchi
design of experimental methodology (DOE)

Roya Khosravi, Seyed Nezamedin Hosseini, Amin Javidanbardan, Maryam


Khatami, Hooman Kaghazian & Seyed Dawood Mousavi Nasab

To cite this article: Roya Khosravi, Seyed Nezamedin Hosseini, Amin Javidanbardan,
Maryam Khatami, Hooman Kaghazian & Seyed Dawood Mousavi Nasab (2019): Optimization
of non-detergent treatment for enveloped virus inactivation using the Taguchi design of
experimental methodology (DOE), Preparative Biochemistry and Biotechnology, DOI:
10.1080/10826068.2019.1599398

To link to this article: https://doi.org/10.1080/10826068.2019.1599398

Published online: 29 Apr 2019.

Submit your article to this journal

Article views: 2

View Crossmark data

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=lpbb20
PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY
https://doi.org/10.1080/10826068.2019.1599398

Optimization of non-detergent treatment for enveloped virus inactivation using


the Taguchi design of experimental methodology (DOE)
Roya Khosravia, Seyed Nezamedin Hosseinia,b , Amin Javidanbardana, Maryam Khatamia, Hooman
Kaghaziana,b, and Seyed Dawood Mousavi Nasaba
a
Department of Recombinant Hepatitis B Vaccine, Production and Research Complex, Pasteur Institute of Iran, Tehran, Iran; bViral Vaccines
Research Center, Pasteur Institute of Iran, Tehran, Iran

ABSTRACT KEYWORDS
In mammalian cell culture technology, viral contamination is one of the main challenges; and, so Optimization; enveloped
far, various strategies have been taken to remove or inactivate viruses in the cell-line production virus inactivation; HSV-1;
process. The suitability and feasibility of each method are determined by different factors includ- Taguchi DOE; CPE
ing effectiveness in target virus inactivation, maintaining recombinant protein stability, easiness—
in terms of the process condition, cost-effectiveness, and eco-friendliness. In this research, Taguchi
design-of-experiments (DOE) methodology was used to optimize a non-detergent viral inactivation
method via considering four factors of temperature, time, pH, and alcohol concentration in an
unbiased (orthogonal) fashion with low influence of nuisance factors. Herpes Simplex Virus-1
(HSV1) and Vero cell-line were used as models for enveloped viruses and cell-line, respectively.
Examining the cytopathic effects (CPE) in different dilutions showed that pH (4), alcohol (15%),
time (120 min), and temperature (25  C) were the optimal points for viral inactivation. Evaluating
the significance of each parameter in the HSV-1 inactivation using Taguchi and ANOVA analyses,
the contributions of pH, alcohol, temperature and time were 56.5%, 19.2%, 12%, and 12%, respect-
ively. Examining the impact of the optimal viral treatment condition on the stability of model
recombinant protein-recombinant human erythropoietin, no destabilization was detected.

Introduction cell banks (endogenous contamination), or contaminated raw


materials and GMP failures (adventitious contamination) in the
There is a continuous rise in the number of approved protein
production and purification processes.[9,13]
therapeutics, and there is little doubt that biopharmaceuticals
In contrast to bacteria and yeast, the detection of viruses
have the potential to become the treatment choices of the is not a straightforward procedure because they are often
future.[1–4] Mammalian cells are one of the most suitable host not observable by routine light microscopy.[9,14,15] Besides,
cells for production of recombinant therapeutic proteins new viruses or certain viruses at low concentrations in the
because of their superiority in terms of proper protein folding, cell culture can escape from detection due to the limitations
assembly and specific post-translational modification and gly- of present testing methods.[1] To ensure the safety of recom-
cosylation compared to other host cells such as bacteria, binant therapeutic protein derived from mammalian cells,
plants, and yeasts.[2,5–7] Hence, for the expression of complex the cell cultures and raw materials must be tested for poten-
recombinant therapeutic proteins such as glycoproteins, tial contaminations (preventive measures).[12,14,16] Moreover,
mammalian cells are preferred to other types of host cells. The in the manufacturing process of the medicinal products,
most commonly used cell lines for the commercial production the respective downstream processes must also have the
of recombinant proteins are Chinese hamster ovary (CHO), capability of clearing viruses from the product stream
baby hamster ovary (BHK) covering 70% of all produced by implementing multiple viral clearance steps (corrective
recombinant therapeutic proteins in the market.[6,8] measures).[10,11,17] For viral clearance validation, both pre-
Although cell culture processes, both recombinant and non- ventive and corrective actions should be conducted accord-
recombinant, seem to be promising technologies for vaccine ing to the ICH Q5A guideline.[10,11,17,18]
production, their susceptibility to contamination by various Virus clearance is typically achieved by inactivation and/
viruses is one of the major challenges.[1,9] Since many of these or removal.[1,8,14] The virus removal procedure is intended
products are injectable, biotherapeutic manufactures must to physically separate viral particles, having high physio-
always ensure the viral safety of the biotechnology prod- chemical resistance, from the intended product.[8,16] The
ucts.[10–12] Viral infection can be introduced from contaminated most widely utilized separation techniques for virus removal

CONTACT Seyed Nezamedin Hosseini Seyyednezam@yahoo.com, hosseini@pasteur.ac.ir Department of Recombinant Hepatitis B Vaccine, Production and
Research Complex, Pasteur Institute of Iran, 3159915111 Tehran, Iran.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/lpbb
ß 2019 Taylor & Francis Group, LLC
2 R. KHOSRAVI ET AL.

are column chromatography, nano-filtration, and precipita- using the outer orthogonal array.[31,33] Analyzing the test
tion.[16,18,19] The virus inactivation as a second approach results in the Taguchi method, we not only can achieve the
is aimed to reduce the infectivity of viruses by chemical optimal condition for the viral inactivation but also we can
and/or physical treatments such as employing detergents assess the role and contribution of each factor separately on
and UV light, as well as incubating infected sample in low- viral inactivation.[31,33]
pH and high-temperature conditions.[10,20–23] The enveloped In this study, using Taguchi DOE methodology, the
viruses, encapsulated by an outer membrane, are generally impact of four factors of temperature, time, pH, and alcohol
more susceptible to solvents, detergents and low pH due to concentration in three levels was examined on viral inacti-
damage or solubilization of their membrane lipids and vation of HSV-1. Based on Taguchi and ANOVA analyses,
proteins.[14,20,24] Solvent-detergent treatment is considered for each parameter, the significance and the optimal value
as a robust method for inactivation of viruses, and it is were identified; finally, the impact of optimal treatment
widely used in industry.[11] However, usage of detergents condition on the stability of model recombinant protein was
has raised environmental concerns since common detergents investigated.
in the biopharmaceutical industry are not considered eco-
friendly.[13] The low-pH treatment is another simple, robust
and effective method, which is typically not detrimental to Materials and methods
product quality, and, unlike detergent inactivation, it does Materials
not introduce components that must be cleared from the
process.[10,23] The low-pH condition usually obtained by Trypsin and fetal bovine serum (FBS) were purchased from
adding a weak acid to the sample. Envelop viral reduction, Gibco Company, and other analytical grade chemicals were
at least in 4 logs, usually requires incubation at pH between purchased from Merck company (Germany). All buffers
3.0 and 4.0 with a typical period of 1 h.[11,22] The selected such as sodium acetate, sodium periodate, carbonate,
ranges for the main factors—including pH, time and tem- and phosphate buffered saline (PBS) were prepared with
perature—in the virus inactivation treatment depends on the analytical grade chemicals purchased from Sigma-Aldrich
stability of the target biological product.[22] Company (Germany), and Water for Injection (WFI) from
For evaluating the efficacy of viral clearance methods Pasteur Institute of Iran. Recombinant human erythropoietin
often a model virus, representing similar infectious viruses (Rec-hEPO) derived from CHO, Vero cell line (Origin –
in terms of structure, is selected.[11,12,25] Herpes virus, rang- African Green monkey (Cercopithecus aethiops)) kidney and
ing from 150 to 200 nm in diameter with large DNA, is one strain Patton of HSV-1 were obtained from the Pasteur
of the mainly used models for enveloped viruses.[24,26,27] Institute of Iran. For cell cultivation, 25 cm2 T-flask and cell
Among various herpes viruses, Herpes Simplex Virus-1 culture 6-well plates, both from BIOFIL (China), and a CO2
(HSV-1) and Herpes Simplex Virus-2 (HSV-2) are the most incubator from New Brunswick (USA) were used. For ana-
common contagious pathogens in humans.[24,26,27] In add- lysis, inverted microscope from Letiz (Germany), ELISA
ition to the model virus, type of cell-line should be specified (Cat. No 11693417001) from Roche (Switzerland) and Size
as well. The Vero cell-line derived from African Green Exclusion High-Performance Liquid Chromatography (SEC-
Monkey Fibroblast Cells is one of the standard cell lines HPLC) with TSK-PW3000 column were employed.[39,40]
which is commonly used for the production of viral vac-
cines.[28–30] These cells grow in a medium adherent to the Methods
bottom of the culture vessel (dependent on the support),
and by reaching the optimum cell density, they form conflu- Cell line cultivation
ent monolayers on the surface.[28] The whole procedure of cell cultivation and viral inactiva-
So far, Optimizing and statistical analyses of non-deter- tion is illustrated in Fig. 1. At first, to thaw Vero cells origi-
gent physicochemical treatments using Taguchi design-of- nating from the WCB, the cryovial was removed from
experiments (DOE) methodology for the inactivation of the nitrogen tank and quickly transferred to a water bath at
enveloped viruses such as HSV-1 has not been reported yet. 37 C as shown in Fig. 1. The vial was continuously and
In general, it can be argued that statistical DOE, is one of gently agitated until the cells were thawed, taking approxi-
the powerful tools for improving product quality and pro- mately 60–90 s. Afterward, the vial content was transferred
cess performance by obtaining optimal production condi- to a centrifuge tube containing 9.0 mL complete culture
tions and presenting valuable statistical model.[31] Using medium (Fig. 1). The culture medium was prepared before
orthogonal arrays, Taguchi’s method is a well-known statis- in 50 ml sterilized tubes with 5 ml of FBS (10% Vol) and
tical method which can be used to examine and predict the 45 ml of DMEM medium. Before adding FBS into the cul-
effect of each factor in empirical studies by the lowest pos- ture media, its bottle was placed in the water bath at 55  C
sible number of experiments with minimum costs.[31,32] water for 20 min to inactivate its complements. Centrifuging
Although the Taguchi method has been used in the elec- at 200 g for 5 min at room temperature, the cells were pel-
tronics and mechanical industry for many years, it has been leted, and the supernatant was discarded. Afterward, as
only recently applied in the fields of biotechnology and shown in Fig. 1, the cells were resuspended by DMEM/FBS
applied biochemistry.[33–38] In this robust design method, culture medium and transferred into a T-flask for prolifer-
the influence of uncontrollable or noise factors is reduced ation (Fig. 1). The adherent Vero cells were grown in the
PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY 3

Figure 1. The schematic diagram of viral inactivation process. In summary, the Vero cells were cultivated and resuspended by trypsinization and added evenly into
6-well plates. The HSV-1 sample at different dilution rates treated using various alcohol concentration, pH and temperature, and time parameters. After the addition
of the treated virus samples into the Vero cells, the Cytopathic effect (CPE) was determined after 72 h’ incubation.

T-flasks inside an incubator at 37  C/30 rpm, 5% CO2. diluting the treated sample with FBS/DMEM culture at dilu-
Reaching cells to above 90% confluent monolayer, they were tion rates of 101, 102, 103, 104, and 105.
resuspended by trypsinization and added evenly into 6-well As illustrated in Fig. 1, by the formation of Vero cells
plates (Fig. 1). After 24 h the color and transparency of the monolayer on plates of six wells after 48 h, the culture media
culture medium were examined; using a reverse microscope, of wells were discarded, and the monolayer was washed out
the cell density and microbial contamination were tested.[41] by phosphate buffer saline (PBS). Then, PBS was discarded,
After detecting Vero cells’ monolayer on all wells, viral and 0.5 ml of treated virus samples in five dilution rates sep-
inactivation procedure was initiated. arately were poured into each well of the plates. Each dilu-
tion plated in duplicate–using two wells–to enhance the
accuracy. In each plate also two wells were chosen as a cell
Viral inactivation blank and a virus blank, respectively. As a result, 12 wells—
We employed HSV-1 stock with the titer of 2108 plaque- two 6-well plates—were used to include all the five dilutions
forming units per milliliter (PFU/ml). As shown Fig. 1, of virus stock sample for each single viral inactivation
100 ml of the solution containing 50 mM of sodium acetate experiment. Following, the plates were incubated at 37  C in
and ethanol (variable) was mixed with 100 ml HSV-1 stock 5% CO2; every 20 min they were removed from the incuba-
in the Eppendorf and then, treated under different time, pH tor and were shaken slowly (Fig. 1). After 1 h, the plates
and temperature conditions designed using Taguchi DOE were removed from the incubator and, after discharging the
methodology. Subsequently, five samples were prepared by culture media, the wells were refilled with 2 ml FBS/DMEM
4 R. KHOSRAVI ET AL.

medium and agar and again placed into the incubator for uncontrollable parameters. Thus, a measurement called sig-
72 h (Fig. 1) to determine the Cytopathic effect (CPE). nal-to-noise ratio (S/N ratio), a measure of signal strength
in respect to the background noise, is used to analyze the
results of experiments in the Taguchi method to omit
Plaque assay using Giemsa stain uncontrollable standard deviation factors. The value of S/N
The plates were removed from the incubator after 72 h, and is calculated based on the following equation.
the culture medium of each well was poured out, and the 
cell surface was washed with PBS solution. Next, as shown
S
N
¼ 10logMSD (1)
in Fig. 1, cells in each well were fixed with 0.5 ml methanol In the above equation, MSD is the mean square of the
and stained with 1 ml of Giemsa stain. Each plate was deviations which is calculated from three different equations
shaken gently so that all the surface of the cells was stained, depending on the objective of the study. Considering that
and after 20 min, the surface of each well was rinsed with this study aims to achieve the conditions in which CPE is
WFI. After air drying, wells were ready for quantifying pla- the minimum and close to zero, the optimal condition—the
ques using the inverted microscope. desired response—is “closer is better.” With this description,
MSD is obtained from Eq. (2).
Pr
DOE for viral inactivation optimization ðYi  MÞ2
The design of experiment by Taguchi method involves the MSD ¼ i¼1 (2)
r
following steps: I. Identifying independent variables and
responses. II. Selection of the number of levels for each
independent variable. III. Selection of suitable orthogonal Testing stability of recombinant protein
array IV. Conducting experiments. V. Analyzing the data Obtaining optimum condition for inactivation of HSV-1, we
and inference. investigated the impact of the optimized treatment on pro-
Using the Taguchi method, four independent parameters tein stability. For this purpose, various concentrations of
of temperature, time, alcohol concentration and pH at three rec-hEPO active pharmaceutical ingredient (rec-hEPO API)
levels were defined for inactivation of the HSV-1 as repre- derived from CHO cell-line (Pasteur Institute of Iran) were
sented in Table 1. The number of plaques, viral titer, was treated under the selected optimal condition for viral inacti-
selected as a response in the designed experiment. Qualitek- vation. HPLC-SEC analysis was performed to examine the
4 software Version 7.0 was used for DOE and analysis of changes in the heterogeneity of the sample. Also, the anti-
Taguchi experiments. Due to the number of parameters and body binding activity of treated rec-hEPO was analyzed by
levels, the Taguchi L9 orthogonal array was selected for EPO ELISA kit (11693417001 Roche).
designing the test. The criteria of all nine experiments are
shown in Table 2.
Results
The Taguchi method is one of advanced statistical quality
improvement techniques. One of the necessary conditions Since no noticeable CPE was observed in the wells in which
for quality improvement is the uniformity of the response the treated virus samples at the dilution rates of 103, 104,
meaning the low noise-variation in its amount due to and 105 were added, the results obtained from the dilution
rates of 0.1 and 0.01 were used as responses of DOE in
Table 1. Values of independent variables at different levels in Taguchi design-
Qualitek-4 software. Fig. 2 demonstrates the responses for
of-experiment methodology. the nine designed experiments, T1–T9 (Table 2), at the dilu-
Levels tion rate of 0.1. The results of the dilution rate of 0.01 are
Process parameters 1 2 3
50
Temperature ( C) 4 25 37 Quantified viral-induced CPE 1
Time (min) 30 60 120
Quantified viral-induced CPE 2
pH (–) 4 5 7
40
Ethanol concentration (v/v, %) 15 10 5
Virus titer (PFU/ml)

30
Table 2. The selected levels for four factors (temperature, time, pH, and etha-
nol) according to an L9 orthogonal array.
20
Values
No. Temp ( C) Time (min) pH (–) Ethanol Conc. (v/v, %)
T1 4 30 4 15 10
T2 4 60 5 10
T3 4 120 7 5
T4 25 30 5 5 0
T5 25 60 7 15 T1 T2 T3 T4 T5 T6 T7 T8 T9
T6 25 120 4 10 Experiment number
T7 37 30 7 10 Figure 2. Virus titer based on quantified viral-induced cytopathic effects in
T8 37 60 4 5 Vero cells in the nine designed experiments using treated virus sample at the
T9 37 120 5 15 dilution rate of 0.1.
PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY 5

Table 3. Mean of standard deviations and signal-to-noise analysis data in presented in appendices (Fig. A1). Since for each experiment
experiments using virus sample at the dilution rate of 0.1.
the treated virus sample at each dilution rate was added
(Y1)2: (Y2)2:
Square of 1st Square of 2nd S/Nb:
identically to two wells, two plaque responses are presented
Treatment run results run results MSDa 10 log MSD in Fig. 2. As illustrated in Fig. 2, no CPE was observed
T1 e62 e62 e62 800 under the selected parameters for experiment T1
T2 25 9 17 12.3045 (Temperature 4  C, Time 30 min, pH: 4, Ethanol concentra-
T3 49 25 37 15.682
T4 25 64 44.5 16.4836 tion: 15%) and experiments T6 (Temperature 25  C, Time
T5 1 9 5 6.9897 120 min, pH: 4, Ethanol concentration: 10%). The same
T6 e62 e62 e62 800 results were obtained for experiments T1 and T6 at the dilu-
T7 9 25 17 12.3045
T8 16 16 16 12.0412 tion rate of 0.01 (Fig. A1).
T9 1 9 5 6.9897 Adding the results of the experiments to the DOE soft-
a
MSD: Mean of standard deviations. ware, the data were systemically analyzed and interpreted
b
S/N: Signal-to-noise ratio. according to the Taguchi method. In this approach, MSD
and subsequently S/N ratio for each treatment were calcu-
lated according to Section “Testing stability of recombinant
Table 4. Main effects of each parameter based on S/N ratio in experiments protein”. The obtained values are presented in Tables 3
using virus sample at the dilution rate of 0.1. and 4. Figure 3 also illustrates the main effects plot for each
Parameter S/NMi,1b S/NMi,2 S/NMi,3 S/N Average level of each factor based on the average S/N ratios pre-

M ¼ Temp, i ¼ 4 C
a a
800 12.3045 15.682 257.3378 sented in the final column of Table 4. The obtained S/N
M ¼ Temp, i ¼ 25  C 16.4836 6.9897 800 258.8422 ratios were also used to form the table of the ANOVA ana-
M ¼ Temp, i ¼ 37  C 12.3045 12.0412 6.9897 10.4451
M ¼ Time, i ¼ 30 min 800 16.4836 12.3045 257.0706 lysis to determine the relative contribution of each factor on
M ¼ Time, i ¼ 60 min 12.3045 6.9897 12.0412 10.4451 viral inactivation (Table 5). The ANOVA table at the dilu-
M ¼ Time, i ¼120 min 15.682 800 6.9897 259.1094 tion rate of 0.01 is presented in appendices (Table A1).
M ¼ pH, i ¼ 4 800 800 12.0412 529.3196
M ¼ pH, i ¼ 5 12.3045 16.4836 6.9897 11.9259 The optimal levels are the points where the target (S/N)
M ¼ pH, i ¼ 7 15.682 6.9897 12.3045 11.6587 ratio is the highest value. According to Fig. 3, the level two
M ¼ Ethanol Conc., i ¼ 15% 6.9897 6.9897
M ¼ Ethanol Conc., i ¼ 10%
800
12.3045 800 12.3045
262.0069
258.4637
of temperature (25  C), the level three of time (120 min), the
M ¼ Ethanol Conc., i ¼ 5% 15.682 16.4836 12.0412 14.7356 level one of the pH (4) and the level one of alcohol concen-
a
M is substituted with each factor (Temperature, Time, pH and Ethanol); i is tration (15%) were the optimal levels. According to the stat-
the chosen level of each factor. istical assessment by ANOVA (Table 5), pH with 56.5%,
b
S/NMi,1, S/NMi,2 and S/NMi,3 indicate the three results of S/N ratio of treat-
ments (T1–T9) which each level of each factor involves in it, for instance, alcohol with 19.2% and temperature, and time with about
M ¼ Time, i ¼ 60 min is involved in T2, T5, and T8 treatments so S/NTime60,1, 12% each contributed to the inactivation of HSV-1. As a
S/NTime60,2, and S/NTime60,3 are S/N ratio of treatment T2, T5, and T8. confirmatory test, the impact of viral inactivation under

Main Effects Plot (data means) for S/N ratio


550

500

450

400

350
Mean of S/N ratio

300

250

200

150

100

50

0
Temp (4 °C) Temp (25 °C) Temp (37 °C) Time (30 min) Time (60 min) Time (120 min) pH (4) pH (5) pH (7) Ethanol Conc. Ethanol Conc. Ethanol Conc.
(15%) (10%) (5%)
-50
Figure 3. Main effects plot of each factor based on the Taguchi Method for the three categorized levels according to the average S/N ratios in experiments using
virus sample at the dilution rate of 0.1.
6 R. KHOSRAVI ET AL.

Table 5. Analysis of Variance (ANOVA) based on the Taguchi method in experiments using virus sample at a dilution rate of 0.1.
Degree of Contribution
Factors freedom Sums of Square Variance (V) F-ratio Pure Sum Percentage (P %)
Temperature 2 1273.659 636.829 0.000 1273.659 12.130
Time 2 1276.499 638.249 0.000 1276.499 12.157
pH 2 5928.681 2964.340 0.000 5928.681 56.466
Ethanol 2 2020.699 1010.349 0.000 20.699 19.245
Error 0 0.000 0.000 0.000
Total 8 10,499.539 100

0.4 to implement suitable alternative treatment condition in


rec-hEPO Before Treatment
rec-hEPO After Treatment
case the protein of interest is unstable on the obtained opti-
0.35
mum value for pH or alcohol. In such situations, the
Absorbance Unit-280 nm (AU)

0.3 remained influential parameters would be adjusted to reach


0.25
viral inactivation based on their significance and effect.
We illustrated that by optimizing the inactivation
0.2
method via Taguchi DOE, it is possible to reduce envel-
0.15 oped viruses effectively at low concentration of alcohol
without using detergent, experiments T1 and T6 (Fig. 2).
0.1
Thus, by the elimination of the usage of detergent in the
0.05 purification process a more economical and environmen-
0
tally friendly production process could be achieved. The
0 50 100 150 200 obtained analysis using Taguchi method indicated that the
Time (Min)
pH played the most significant role in viral inactivation.
Figure 4. The SEC-HPLC analysis of Active Pharmaceutical Ingredient (API) sam- The main reason for the importance of this parameter is
ple of rec-hEPO before and after virus inactivation treatment. In this chromato-
gram, the main peak represents the rec-hEPO monomers. The small peaks that lipid-enveloped viruses are susceptible to low pH,
before the main peak belong to the oligomers and dimers of rec-hEPO, and they are denatured in acidic conditions.[19,23,24] The
respectively. obtained conclusion regarding the impact of low pH on
viral inactivation confirms previous studies which reported
optimal condition was reevaluated which supported our pre- acidic pH (3.9 or below) incubation as an effective tool for
vious observation meaning complete HSV-1 inactivation. inactivation of retroviruses.[11,12,19,24] In this study, the
Achieving the optimal virus inactivation condition, rec- minimum pH was selected to the value of 4.0 (Table 1), as
hEPO protein sample was also treated under the optimized several proteins are unstable in the lower values. Higher
condition to examine the stability of the recombinant pH values were also not selected in this study since accord-
protein. According to HPLC-SEC analysis (Fig. 4), the ratio ing to the literature, in contrast to low pH, incubation in
of monomers (the main peak area) to total peak area basic condition, e.g., pH 9, had shown no significant effect
remained almost constant meaning that the heterogeneity on inactivation of viruses.[42]
of the sample was not changed. ELISA assay also supported Based on the analysis by the Taguchi method, ANOVA
that the antigenicity of the sample was not modified after table (Table 5), among other three remained parameters
the treatment. ethanol concentration demonstrated the greatest significance
compared to time and temperature. It was previously
Discussion reported that enveloped viruses have very low resistance to
20% ethanol.[14,43] In this study, 15% v/v ethanol was
Viral contamination can be considered as one of the major selected as the highest concentration level, since there is
issues in the biopharmaceutical industry, and if it is not always a possibility of denaturation and aggregation of
detected and eliminated, in addition to biosafety issue the therapeutic proteins at higher concentrations.[44,45] Selection
reputation of the company will also be adversely affected. of low concentrations of ethanol in this study could be the
Due to the versatility of viruses as well as recombinant main reason for its notably lower significance, 19.2%, com-
products, there is no holistic approach which can provide pared to pH, 56.5%, in viral inactivation based on ANOVA
complete viral inactivation in biotechnology products. In (Table 5). Non-the-less, the obtained results suggest that
this study, to minimize the changes in the composition of when the recombinant protein is susceptible to low pH but
the solution, we did not use detergent. Instead, using a more resistant to high ethanol concentration, a higher con-
robust statistical tool, pH and ethanol, temperature, and centration of alcohol could replace the pH effect on viral
time were examined since these parameters could easily be inactivation.
employed and controlled in the production process and also Main effects based on Taguchi method also indicated that
do not add any significant burden on purification process as parameters of temperature and time demonstrated little
well as wastewater treatment systems. The effect and signifi- effects on viral inactivation (Fig. 3). Although dry heat treat-
cance of each of four parameters on the HSV-1 inactivation ment in temperatures above 60  C demonstrated the lethal
were examined by DOE methodology. Obtaining significance effect on viruses especially HSV,[16,25] in temperatures below
of each factor on viral inactivation provides the possibility this value—the selected range in this study—there was no
PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY 7

considerable impact on clearance of viruses.[26] Similar to Disclosure statement


the ethanol the main reason for the defined temperature
No potential conflict of interest was reported by the authors.
range was that the most recombinant therapeutic proteins
are susceptible to higher temperature conditions.[26,42]
Besides, a substantial increase and a subsequent decrease in ORCID
the temperature is not favorable in production plants due to Seyed Nezamedin Hosseini http://orcid.org/0000-0002-1457-0974
related extra costs. Nevertheless, the temperature was
selected as a factor with the maximum value of 37  C to
investigate its impact on viral inactivation under the non-
favorable condition for the virus due to pH and alcohol References
parameters. Regarding the degree of significance of time as a [1] Darling, A. Validation of Biopharmaceutical Purification
parameter, based on the results we can conclude that the Processes for Virus Clearance Evaluation. MB. 2002, 21,
major portion of viral inactivation occurred via pH and 057–084.
[2] Bielser, J.M.; Wolf, M.; Souquet, J.; Broly, H.; Morbidelli, M.
ethanol in a relatively short period—about 30 min; thus,
Perfusion Mammalian Cell Culture for Recombinant Protein
increasing the time length did not show any significant Manufacturing – A Critical Review. Biotechnol. Adv. 2018, 36,
effect on the response (Fig. 3 and Table 5).[24] 1328–1340.
Regarding recombinant protein stability, it was previ- [3] Nezamedin Hosseini, S.; Javidanbardan, A.; Sadat Alizadeh
ously reported that using a high concentration of recom- Salim, B.; Khatami, M. Large-Scale Purification of Recombinant
binant protein increases its stability against treatments such Hepatitis B Surface Antigen from Pichia pastoris with Non-
Affinity Chromatographic Methods as a Substitute to
as low pH incubation.[46] Besides previous studies claimed Immunoaffinity Chromatography Large-Scale Purification of
that increasing recombinant protein concentration possibly Reco. Prep. Biochem. Biotechnol. 2018, 48, 683–692.
increase the efficacy of viral inactivation.[10,13,47] In this [4] Kimia, Z.; Hosseini, S.N.; Ashraf Talesh, S.S.; Khatami, M.;
study, we did not statistically evaluate the impact of recom- Kavianpour, A.; Javidanbardan, A. A Novel Application of Ion
binant protein concentration on viral inactivation using Exchange Chromatography in Recombinant Hepatitis B
Vaccine Downstream Processing: Improving Recombinant
DOE. However, in the final stage after confirming DOE
HBsAg Homogeneity by Removing Associated Aggregates.
methodology, the impact of the optimal condition on the J. Chromatogr. B 2019, 1113, 20–29.
stability of rec-hEPO model protein with different concen- [5] Hesse, F.; Wagner, R. Developments and Improvements in the
trations was examined by HPLC-SEC and ELISA analyses. Manufacturing of Human Therapeutics with Mammalian Cell
According to the results, no significant change in the anti- Cultures. Trends Biotechnol. 2000, 18, 173–180.
[6] Wurm, F.M. Production of Recombinant Protein Therapeutics
genicity and heterogeneity of the rec-hEPO was detected. in Cultivated Mammalian Cells. Nat. Biotechnol. 2004, 22,
Perez et al. in their study also reported that no change in 1393–1398.
the biological activity of rec-hEPO was detected by employ- [7] Gazor, M.A.; Talesh, S.S.; Hosseini, S.N.; Javidanbardan, A.;
ing viral inactivation method with 50% ethanol.[14] The Khatami, M. High Recovery of Intracellular Recombinant
results of our study suggest that viral inactivation with HBsAg from Pichia pastoris via Continuous Pulsed Laser Cell
Disruption System Optimized by Response Surface
significantly lower ethanol percentage—15% v/v—is still
Methodology. Biotechnol. Appl. Biochem. 2019, 66, 91–100.
effective, without any harsh impact, such as denaturation [8] Asper, M.; Hanrieder, T.; Quellmalz, A.; Mihranyan, A.
and aggregation, on the structure and biological activity of Removal of Xenotropic Murine Leukemia Virus by
the target recombinant protein (Fig. 4). Since in this study Nanocellulose Based Filter Paper. Biologicals 2015, 43, 452–456.
compared to the study of Perez et al., much lower ethanol [9] Merten, O.W. Virus Contaminations of Cell Cultures – A
Biotechnological View. Cytotechnology 2002, 39, 91–116.
concentration was employed for viral inactivation, it seems [10] Chinniah, S.; Hinckley, P.; Connell-Crowley, L. Characterization
to be reasonable to use this optimized treatment method of Operating Parameters for XMuLV Inactivation by Low pH
for other fragile and less stable recombinant proteins in Treatment. Biotechnol. Progress 2016, 32, 89–97.
high alcohol concentration. [11] Shukla, A.A.; Aranha, H. Viral Clearance for Biopharmaceutical
Downstream Processes. Pharm. Bioprocess 2015, 3, 127–138.
[12] Ellgaard, T.W.; Bindslev, L.; Kamstrup, S. Evaluation of the
Conclusion Virus Clearance Capacity and Robustness of the Manufacturing
Process for the Recombinant Factor VIII Protein, Turoctocog
By implementing Taguchi DOE methodology, we optimized alfa. Protein Expr. Purif. 2017, 129, 94–100.
non-detergent treatment for inactivation of enveloped virus, [13] Conley, L.; Tao, Y.; Henry, A.; Koepf, E.; Cecchini, D.; Pieracci,
J.; Ghose, S. Evaluation of Eco-Friendly Zwitterionic Detergents
HSV-1 with the low influence of uncontrollable factors.
for Enveloped Virus Inactivation. Biotechnol. Bioeng. 2017, 114,
Based on Taguchi and ANOVA analyses, the significance of 813–820.
each parameter was also identified. The obtained results [14] Perez, M.; Rodrıguez, E.; Rodrıguez, M.; Paez, R.; Ruibal, I.;
indicated that under the optimum condition effective viral Noa, E.; Garcıa, O.; Moya, G.; Martınez, M.; Marcelo, J.; et al.
inactivation could be achieved without affecting the stability Validation of Model Virus Removal and Inactivation Capacity
of the target recombinant protein, in this case, rec-hEPO. of an Erythropoietin Purification Process. Biologicals 2011, 39,
430–437.
This technique is eco-friendly, straightforward and rational [15] Ali, S. Microbial and Viral Contamination of Animal and
viral inactivation approach which can be easily implemented Stem Cell Cultures: Common Contaminants, Detection and
in cell culture processes. Elimination. J. Stem Cell Res. Ther. 2017, 2, 00078.
8 R. KHOSRAVI ET AL.

[16] Valdes, R.; Ibarra, N.; Ruibal, I.; Beldarraın, A.; Noa, E.; Binding Capacity of Anion Exchange Chromatography Media
Herrera, N.; Aleman, R.; Padilla, S.; Garcia, J.; Perez, M.; et al. for Recombinant Erythropoietin Purification. Iran. J. Biotechnol
Chromatographic Removal Combined with Heat, Acid and 2014, 12, 49–55.
Chaotropic Inactivation of Four Model Viruses. J. Biotechnol. [33] Rao, R.S.; Kumar, C.G.; Prakasham, R.S.; Hobbs, P.J. The
2002, 96, 251–258. Taguchi Methodology as a Statistical Tool for Biotechnological
[17] Rathore, A.S.; Kumar, V.; Arora, A.; Lute, S.; Brorson, K.; Applications: A Critical Appraisal. Biotechnol. J. 2008, 3,
Shukla, A. Mechanistic Modeling of Viral Filtration. J. Memb. 510–523.
Sci. 2014, 458, 96–103. [34] Prasad, K.K.; Mohan, S.V.; Rao, R.S.; Pati, B.R.; Sarma, P.N.
[18] Etzel, M.R.; Riordan, W.T. Viral Clearance Using Monoliths. Laccase Production by Pleurotus ostreatus 1804: Optimization of
J. Chromatogr. A 2009, 1216, 2621–2624. Submerged Culture Conditions by Taguchi DOE Methodology.
[19] Roberts, P.L.; Dunkerley, C.; Walker, C. Virus Reduction Biochem. Eng. J. 2005, 24, 17–26.
in an Intravenous Immunoglobulin by Solvent/Detergent [35] Nandal, P.; Ravella, S.R.; Kuhad, R.C. Laccase Production
Treatment, Ion-Exchange Chromatography and Terminal Low by Coriolopsis caperata RCK2011: Optimization under Solid
pH Incubation. Biologicals 2012, 40, 345–352. State Fermentation by Taguchi DOE Methodology. Sci. Rep
[20] Kingsley, D.H.; Chen, H. Influence of pH, Salt, and 2013, 3, 1386.
Temperature on Pressure Inactivation of Hepatitis a Virus. [36] Chenthamarakshan, A.; Parambayil, N.; Miziriya, N.; Soumya,
Int. J. Food Microbiol. 2009, 130, 61–64. P.S.; Lakshmi, M.S.K.; Ramgopal, A.; Dileep, A.; Nambisan, P.
[21] Geisler, C.; Jarvis, D.L. Adventitious Viruses in Insect Cell Optimization of laccase production from Marasmiellus palmivo-
Lines Used for Recombinant Protein Expression. Protein rus LA1 by Taguchi method of Design of experiments. BMC
Expression and Purification 2018, 144, 25–32. Biotechnol. 2017, 17, 12.
[22] Orozco, R.; Godfrey, S.; Coffman, J.; Amarikwa, L.; Parker, S.; [37] Periasamy, R.; Palvannan, T. Optimization of Laccase
Hernandez, L.; Wachuku, C.; Mai, B.; Song, B.; Hoskatti, S.; Production by Pleurotus ostreatus IMI 395545 Using the
et al. Design, Construction, and Optimization of a Novel, Taguchi DOE Methodology. J. Basic Microbiol. 2010, 50,
Modular, and Scalable Incubation Chamber for Continuous 548–556.
Viral Inactivation. Biotechnol. Progress 2017, 33, 954–965. [38] Basak, B.; Bhunia, B.; Mukherjee, S.; Dey, A. Optimization of
DOI: 10.1002/btpr.2442.
Physicochemical Parameters for Phenol Biodegradation by
[23] Ma, R.; Cui, X. Low pH Inactivation for Xenotropic Gamma
Candida tropicalis PHB5 Using Taguchi Methodology. Desalin.
Retrovirus in Recombinant Human TNF-a Receptor
Water Treat. 2013, 51, 34–36.
Immunoglobulin G and Mechanism of Inactivation. Biologicals
[39] Mostafaei, M.; Hosseini, S.N.; Khatami, M.; Javidanbardan, A.;
2014, 42, 52–56.
Sepahy, A.A.; Asadi, E. Isolation of Recombinant Hepatitis B
[24] Weed, D.J.; Pritchard, S.M.; Gonzalez, F.; Aguilar, H.C.; A. V,
Surface Antigen with Antibody-Conjugated Superparamagnetic
N. Mildly Acidic pH Triggers an Irreversible Conformational
Fe3O4/SiO2 core-Shell Nanoparticles. Protein Expr. Purif. 2018,
Change in the Fusion Domain of Herpes Simplex Virus 1
145, 1.
Glycoprotein B and Inactivation of Viral Entry. J. Virol. 2017,
[40] Hosseini, S.N.; Ghaisari, P.; Sharifnia, S.; Khatami, M.;
91, e02123–e02116.
[25] Nims, R.W.; Zhou, S.S. Intra-Family Differences in Efficacy of Javidanbardan, A. Improving the Recovery of Clarification
Inactivation of Small, Non-Enveloped Viruses. Biologicals 2016, Process of Recombinant Hepatitis B Surface Antigen in Large-
44, 456–462. Scale by Optimizing Adsorption-Desorption Parameters on
[26] Lancz, G.J. Physical Integrity of Herpes Simplex Virus Aerosil-380. Prep. Biochem. Biotechnol 2018, 48, 490–497.
Following Thermal Inactivation. Arch. Virol. 1980, 64, 375–381. [41] Davis, J.M. Basic Cell Culture : A Practical Approach. Oxford
[27] Asculai, S.S.; Weis, M.T.; Rancourt, M.W.; Kupferberg, A.B. University Press: Oxford, 2002; 381 pp.
Inactivation of Herpes Simplex Viruses by Nonionic [42] Croughant, W.S.; Behbehani, A.M. Comparative Study of
Surfactants. Antimicrob. Agents Chemother. 1978, 13, 686–690. Inactivation of Herpes Simplex Virus Types 1 and 2 by
[28] Ammerman, N.C.; Beier-Sexton, M.; Azad, A.F. Growth and Commonly Used Antiseptic Agents. J. Clin. Microbiol. 1988,
Maintenance of Vero Cell Lines. Curr. Protoc. Microbiol. 2008, 26, 213–215.
Appendix 4, Appendix 4E. [43] Roberts, P.L.; Lloyd, D. Virus Inactivation by Protein
[29] Sagripanti, J.L.; Routson, L.B.; Bonifacino, A.C.; Lytle, C.D. Denaturants Used in Affinity Chromatography. Biologicals
Mechanism of Copper-Mediated Inactivation of Herpes Simplex 2007, 35, 343–347.
Virus. Antimicrob. Agents Chemother. 1997, 41, 812–817. DOI: [44] Bull, H.B.; Breese, K. Interaction of Alcohols with
10.1128/AAC.41.4.812. Proteins. Biopolymers 1978, 17, 2121–2131.
[30] Frazatti-Gallina, N.M. Purified Vero-Cell Rabies Vaccine. In [45] Nikolaidis, A.; Andreadis, M.; Moschakis, T. Effect of Heat, pH,
Current Laboratory Techniques in Rabies Diagnosis, Research Ultrasonication and Ethanol on the Denaturation of Whey
and Prevention; Rupprecht, C., Nagarajan, T., Eds; Vol. 2. Protein Isolate Using a Newly Developed Approach in the
Academic Press: New York, 2015; pp. 261–268. Analysis of difference-UV Spectra. Food Chem. 2017, 232,
[31] Mojarrad Moghanloo, G.M.; Khatami, M.; Javidanbardan, A.; 425–433.
Hosseini, S.N. Enhancing Recovery of Recombinant Hepatitis B [46] Aranha, H.; Forbes, S. Viral Clearance Strategies for
Surface Antigen in Lab-Scale and Large-Scale Anion-Exchange Biopharmaceutical Safety Part II : A Multifaceted Approach to
Chromatography by Optimizing the Conductivity of Buffers. Process Validation. BioPharm Int. 2001, 14, 43–54.
Protein Expr. Purif. 2018, 141, 25–31. [47] Chen, Q. Viral Clearance Using Traditional, Well-Understood
[32] Sepahi, M.; Kaghazian, H.; Sereshkeh, M.P.; Sadeghcheh, T.; Unit Operations (Session I): Low-pH Inactivation. PDA J.
Hadadian, S.; Jebeli, M.R.; Yavari, F. Optimization of Dynamic Pharm. Sci. Technol. 2014, 68, 17–22.
PREPARATIVE BIOCHEMISTRY AND BIOTECHNOLOGY 9

Appendices

50
Quantified viral-induced CPE 1
Quantified viral-induced CPE 2
40

Virus titer (CFU/ml)


30

20

10

0
T1 T2 T3 T4 T5 T6 T7 T8 T9
Experiment number
Figure A1. Quantified viral-induced cytopathic effects in experiments using virus sample at a dilution rate of 0.01.

Table A2. Analysis of Variance (ANOVA) based on Taguchi method in experiments using virus sample at a dilution rate of 0.01.
Degree of Contribution
Factors freedom Sums of square Variance (V) F-ratio Pure sum percentage (P%)
Temperature 2 1373.208 686.604 0.000 1373.208 15.494
Time 2 950.641 475.320 0.000 950.641 10.726
pH 2 4778.974 2389.487 0.000 4778.974 53.923
Ethanol % 2 1759.756 879.878 0.000 1759.756 19.856
Error 0 0.000 0.000 0.000
Total 8 8862.582 100

Potrebbero piacerti anche